Prato CIRN-DIAC Community Informatics Conference 2010: PhD Stream

Developing a Technology Adoption Model for a Sustainable Agriculture Sector Sinead O’Neill School of Computer Science and Statistics Trinity College, Dublin, Ireland

Abstract

The adoption and use of Information and Communications Technology (ICT) by traditional businesses, which can help deliver benefits is well researched. However, there are parts of society, such as, the agricultural community that have yet to fully explore and investigate the potential benefits of ICT. In Ireland, the agricultural sector account for 8% of Gross Domestic Product and it is an integral part of |Irish society. The effective use of ICTs by Irish farmers can improve their operation and support them to achieve their objectives, for example, access to information and services, social inclusion, cultural provision, education and training. However, the use of ICTs and its associated services by Irish farmers has been limited and often underutilised. The adoption and diffusion of ICT among Irish farmers needs to be approached in a holistic and sustainable fashion that will support their operation and enhance their productivity. This research will develop a conceptual model that will assist Irish farmers adopt and use ICT effectively. The model should enable Irish farmers to use applications that will enhance their operation.

Keywords: Farmers, Agricultural community, ICT adoption and diffusion

Introduction The land area of Ireland is 6.9million hectares where sixty percent is used for agriculture. The agriculture sector can be classified into two divisions the primary agriculture (farming) sector and the agri-business sector. Primary agriculture is the foundation and platform for the prosperity and development of the agriculture and food sector in Ireland. Agri-business is a generic term that refers to the various businesses linked to agriculture such as seed supply, agrichemicals, farm machinery, food production, processing, marketing and retail sales. This sector accounts for 8.5% of employment, 10% of total exports and accounts for 8% of GDP within Ireland (DAFF 2009) making it an important indigenous industry.

Profile of Irish farming According to the Farm Structure Survey 2007 there are 128,200 farms in Ireland (CSO 2008). Farming has strong traditions and ties with the Irish nation and has influenced the landscape of rural Ireland for centuries. Many Irish authors have been influenced by life on the farm. Thomas Kavanagh, John McGahern capture life on the farm through its physical labour in inhospitable climatic conditions, which is always changing according to seasonal changes. Farming activities have more or less stayed the same; milk is brought to the creamery, animals need feeding and the soil is still ploughed in spring to plant crops. Farming has always been viewed as labour intensive which is influenced by many elements such as weather, disease, markets and since Ireland’s membership of the

1

Prato CIRN-DIAC Community Informatics Conference 2010: PhD Stream European Union (EU), the Common Agricultural Policy or CAP1. Ireland’s membership of the EU poses many challenges to agriculture. So how does EU policy effect agriculture? Agriculture within the European Union

Agriculture in the EU is expected to fulfil many functions, the main one being the supply of safe and high quality food to the citizens of Europe. Other functions include maintaining the countryside, helping rural areas remain attractive and competitive places to live. Agriculture is under going fundamental change that requires famers to adapt to new working conditions and seize new market opportunities. All of which are being impacted by climate change. To promote a sustainable sector and a fair standard of living for rural communities the agricultural community will have to explore new business opportunities. These can be found in knowing what consumers’ expectations of agriculture are. Consumers’ expectations of agriculture can be divided into the following: safe food, landscape management, preservation of rural communities and global markets.

Consumers expectations of agriculture Safe Food production

The primary function of agriculture is the productions of safe food. Consumers expect safe food that is competitively priced. Where animals are part of the food chain consumers expect food production that insures the welfare of animals. Ireland along with other EU states has faced obstacles in the production of food from the Foot in Mouth outbreak in 2001 to extreme weather conditions all of which impose strain on the sector. The ability to respond to adversities depends on a dynamic farm sector that can maintain food production in tough conditions and still cater for changing consumer demands. (Riordan 2005) Landscape management

The word farming is synonymous with the word rural a view held by Pyysiainen et al (2006) who suggests “the classic figure of the rural landscape is the farmer”. Agriculture is the main use of land and farming is responsible for shaping and maintaining the rural landscape and eco-system of Europe. The disappearance of farming would result in loosing the environmental and cultural aspects of rural society thus making them bleak places to live. Hoggart et al (1995) noted that the most striking feature about agriculture in Europe is its relative decline. Areas where farmers have left the land have been hit by social and economic pressures, which in turn have made them unattractive places to live. Older generations have been left in isolation while the younger generations have sought a better life in urban areas. Hence, the presence of farmers on the land is a necessary part of the social and economic rural landscape. Preservation of Rural community

Agriculture is the core element of rural life and influences the economic, environmental, and social environment of rural areas. Pyysiainen et al (2006) suggested that agriculture provides the platform for economic diversification and viability of rural areas. Exploring new means for generating income both the farmer and the farm family has a positive effect on the rural community by creating new employment opportunities in rural areas. Average rural incomes are below urban incomes across the EU. Within Ireland average farm incomes fall below the average industrial wage (IFA Conference, 2010). Other avenues for income will have to be explored in order to support rural communities. If not they will remain unattractive to young people and the vibrancy of the rural society will be depleted. Some academics refer to community preservation as ‘social capital’. Putnam (1993, 1995a, 1

CAP – Common Agricultural Policy is a system of European Union agricultural subsidies and programmes. It combines a direct subsidy

payment for crops and land which may be cultivated with price support mechanism. The aim of the common agricultural policy (CAP) is to provide farmers with a reasonable standard of living, consumers with quality food at fair prices and to preserve rural heritage.

2

Prato CIRN-DIAC Community Informatics Conference 2010: PhD Stream 1995b, 2000) popularised the term social capital and defined it as the networks, norms and trust, that facilitate action and cooperation for mutual benefit (Putnam 1993). The central premise behind social capital as noted by Putnam is that ‘social networks have value’ and in the same way that physical or human capital can increase productivity, so too can social networks affect the productivity of groups or individuals. Putnam (2000) outlined a variety of benefits which accrued to communities characterised by having strong social networks. Firstly, he asserted that social norms and the networks that enforce them provide a mechanism whereby individuals can solve collective problems more easily. Secondly, he described how social capital greased the wheels that allow communities to advance smoothly and thirdly social networks can often facilitate the achievement of personal goals as, for example, employment opportunities could be enhanced through personal connections. Finally, he described how social capital also operates through psychological and biological processes to improve individuals’ lives. The basic hypothesis here is that feeling part of an area rich in social capital can have a beneficial effect on the psycho-social conditions of individuals’ lives. The belief held by Putnam (2000) emphasised the importance of farming to rural communities in terms of self preservation or social capital. Global markets

Open markets and globalisation places competitive pressure on farmers (Pitts 2005). The need for agriculture and the extended food sector to exploit new markets with quality products creates its own challenge. Further liberalisation of trade laws will create volatility within the open market, which will results in cheaper food for the consumers. However, the food sector will have to manage these markets shifts while still allowing farmers to generate an income. Constant restructuring and innovative food production will enhance the opportunities within new markets. Farmers will have to adapt to these challenges while at the same time producing quality and safe food. Climate change

Agriculture is vulnerable to climatic changes. Extreme weather conditions have increased the severity of flooding within Irish rural areas. Flooding is happening in both summer and winter months imposing losses on farmers in terms of foodstuff and animals. Areas that are being exposed to extremes in weather conditions will be under threat with continual changes in climate. These weather extremes influence the food markets by making them volatile. However, farming will have to explore the positive side of climate change and one new avenue could be the production of bio-energy products. As Faaij (2004) suggest bio-energy 2 is one of the key areas to mitigate greenhouse gas emissions which is influencing climate change. Changes in mind set among farmers will allow them to consider other crops to grow such as biomass3 crops. These crops have the potential to open up new markets and create new business opportunities for farming. With the increasing demands on agriculture developing a sustainable agriculture sector will depend on developing the famer’s capacity for adapting to new markets situations. This in turn will allow farmers to seize opportunities within new economies. The use of ICT within farm enterprises will have a positive effect for farming in terms of available up to date information. With volatile markets, changing weather conditions and the potential of diseases farmers needs access to current information that will allow them to make business decisions. ICT provides one such avenue for adapting farmers to change. With the changing landscape of farming, farmers will have to explore the use of ICT as a business tool within their farming enterprise. Attempts to understand and improve the effectiveness of ICT adoption for agriculture and rural development remains a concern not only for farmers and academic researchers but also for EU citizens as a whole (Gelb,2008). In an attempt to address this it is important to answer the following research questions. 2

Bio-energy refers to energy produced from biomass. Biomass comprises of residues from forestry and agriculture and specially cultivated crops for energy production such as rapeseed and perennial crops as Willow and Miscanthus (Faaij,2004) 3

3

Prato CIRN-DIAC Community Informatics Conference 2010: PhD Stream

Research Questions Q1. How are Irish farmers currently using Information and Communication Technology (ICT) within their farming enterprises ? Q2. What are the barriers and success factors that affect the adoption and use of ICT by Irish farmers? Q3. What are the concepts and factors that are specific to an Irish farming technology adoption model and how do they interact to provide a useful, usable and sustainable technology adoption model for farmers ? Answers to these questions will determine the level of ICT adoption among famers which in turn will give rise to the development of a conceptual model for ICT adoption among Irish farmers and rural communities. Literature Review Adoption Research

Explaining user behaviour and acceptance of new technologies is often described as one of the more mature research areas within information systems literature (Hu et al 1999;Venkatesh et al 2003). While findings from this research have had an impact on the business world, such as online learning (Saade et al. 2003); internet banking (Lai et al. 2005); technology-related decision-making (Benamati and Rajkumar 2002); other non-traditional business environments such as farming have not been exposed to these findings. The resultant research gap has led to a void in IS literature which is scant in predictive models of farmers behaviour towards ICT. The problem statement of this research therefore is to understand why the farming sector have been slow to adapt ICT technologies such as farming software, eservices or web-based applications (Murphy 2009). Against this background the primary agriculture sector has been quick to adopt new technologies, that involve biological innovations (new seed variations), chemical innovations (fertilizers and pesticides) animal innovations (feeding and breeding) or mechanical technology (tractors and combines). New technologies within these areas offer farmers opportunities to increase production and income and hence are more likely to be adopted (Feder et al 1985). Adoption Models

The basic theoretical question of what are the determinants of the adoption of an innovation needs to be defined. To investigate this question requires the development of a general theory of adoption. However, Rose and Bosie (1974) as citied by Downs and Mohr (1976) rejected the existence of a “unitary theory of innovation” and postulated that distinct types of innovations could be explained by distinct theories. Therefore, a single theory and a set of determinants could be applied to all innovation research should be disbelieved by innovation researchers. Winters (1968) agreed that an innovation was “rarely the same” within two organisations. To understand adoption required an understanding of innovation, as Downs and Mohrs(1976) believed that innovation and adoptability were “two sides of the same coin”. To arrive at a general theory of innovation Downs and Mohr (1976) investigated how a general theory of innovation could be conceptualised. They suggested a set of proposals for innovation research. Conceptual issues raised by them included developing typologies of innovation that could be achieved by dividing attributes of innovation into primary and secondary attributes. Primary attributes were essential to the innovation and could have a constant impact on the innovation an example being cost. Secondary attributes “are perceived by the senses” and could be different dependant on how they are perceived by the person. Downs and Mohr believed that secondary attributes should be measured with respect to each organisation. Interactive models could be used to develop a general theory as a unitary theory does not exist that can be generalised from one organisation to the next. Finally, they also agreed that in studying adoptability of innovations two approached should be made. The first approach is to study the innovations in relation to one organisation or to use the innovation-decision design.

4

Prato CIRN-DIAC Community Informatics Conference 2010: PhD Stream Mathieson (1991) also shared this few and believed that in studying information systems that the “issues that influence use of information systems vary between the system, the individual and the context”. He also added that if these issues could be identified, they could be incorporated into both future systems development and adoption models. To investigate the adoption of technology by the farming community a study of different aspects of technology adoption models was undertaken. Each model was considered under “degree of generality and social variables” as defined by Mathieson (1991). This view was chosen because according to Laumer,Eckhardt and Trunk (2008) adoption decisions varied with “age, social environment and the level of education” which encapsulated Mathiesons “social variables”. However, Gelb and Parker (2005) found that within an agricultural setting, farmer’s performance expectancy was a driver to behavioural intention despite age, social setting, and environment, which stressed the need for the model to be generalizable. The chosen models were the Theory of Planned Behaviour, the Technology Adoption Model, and the Model of Technology Adoption within the Household. 1.

Theory of Planned Behaviour (TPB)

TPB was posited by Fishbein and Ajzen (1975) as a well-researched social science intention model that was used to predict human behaviour within any research environment. Certain actions that were carried out by a person could be executed with a minimal thought process and were often referred to as routine actions. The behaviour executed by a person within these routines could be difficult to determine as the perceived senses of the person towards the object were carried out with a minimal thought process. To overcome this obstacle Ajzen and Fishbein (1975;2002) proposed that behaviour was guided by three general beliefs. Firstly, behavioural belief, this was the belief about the consequence of the behaviour. Secondly, normative beliefs, which were the beliefs a person had about their social norms and thirdly, control beliefs, the presence of factors that may facilitate or impede the behaviour. The intention to behave in a certain manner was based upon these beliefs and in addition, these beliefs were influenced by both external and internal influences thus making them difficult for the researcher to capture. Often a person’s overt action was different from their verbal attitude. In an attempt to model human behaviour, they proposed aggregating the beliefs that could be applied within a research environment to predict a person’s intention to behave (see Figure 1.1). Ajzen and Fishbien (1985) suggested that attitude (A) towards behaviour (B), the subjective norm (SN) and perception of behavioural control (PBC) led to the formation of behavioural intention (BI) which was summarised in the relation: Behavioural Intention BI = (A + SN+ PBC) Figure 1.1 Theory of Planned Behaviour The favourable the attitude (A) and subjective norm (SN), the greater was the perceived control (PBC) and the greater the intention of the person to carry out the behaviour. Intention was assumed the antecedent of behaviour and with a given degree of control over behaviour; a person could carry out the predicted behaviour when the opportunity arose. The attitude a person had towards an object could comprise of many attributes all of which had to be evaluated by the person. Since the overall attitude towards the object was the sum of sub-attitudes to different attributes, it became essential to identify each attributes. This went beyond taking a simple notion of hypnotising an overall attitude to an object. Thus, it was imperative for the researcher to identify the attitudes that were important for the behaviour in question. To achieve this five to nine beliefs were elicited from the population sample through interviews and were referred to as the modal salient beliefs. Degree of generality The capture of attitudes was a successful outcome of TPB, through the adaption of the model to different research environment (Sheppard et al., Madden et al. 1992; Ajzen and Fishbein 1980). The model identified beliefs that were specific to each situation but Ajzen and Fishbein (1980) did not advocate that beliefs that apply in one context could apply to other contexts. However they believed that while some beliefs may generalise across the contexts others may not a view which was also held

5

Prato CIRN-DIAC Community Informatics Conference 2010: PhD Stream by Mathieson (1991). This belief supports Downs and Mohr (1976) view for the need of a non-unitary model of adoption. Conversely, some researchers (Sheppard, Hartwick, & Warshaw 1988) argued that the lack of parsimony of TPB was a limitation of the model. This belief was also held by Ogden (2003) who believed that the TPB constructs where “too general to permit precise tests” and therefore the theory was unable to be rejected. Falsification of the theory did not arise, as one of the model’s predictors failed to carry weight in predicting the intention or behaviour. Ajzen and Fishbein (2003) refuted this by arguing that extra theoretical considerations was required in prediction of behaviour and said that the three antecedents were sufficient. Mathieson (1995) suggested that the TPB requires “a pilot study to identify relevant outcomes, referent groups, and control variables in each context “ were it was used which allowed it to be adapted for each context. Ogden (2003) second argument in refute of the model was that variance in intentions and behaviour was left unaccounted. She was concerned with the fact that in many studies she had reviewed, self-reported rather than objective measurements had been used to measure behaviour. Selfreports of behaviour could in her opinion be contaminated by self-reported cognitions as it led to bias. Armitage and Conner (2001) agreed that bias inflated observed behaviour between cognitive behaviour and actual behaviour. Mathieson (1995) argued that to offset bias TPB required an “explicit behavioural alternative” which allowed respondents to make a comparison within the behaviour been tested. Social variables The presence of social variables (SN) within the model allowed unique variances in intention to be captured. For example, some individuals might use ICT because they were perceived by their peers as being more technologically skilled, which in turn influenced intention to behave. With the integration of beliefs and values into the theory of attitude formation, there was an implicit recognition that although attitudes were themselves permanent; they could change in the face of new knowledge. The second model to be reviewed is the Technology Acceptance Model (TAM) 2. Technology Acceptance Model The Technology Acceptance Model (TAM) was posited by Davis (1986) to determine the intention to adopt a technology by identifying a small number of fundamental variables. The Theory of Reasoned Action (1975) was used as the theoretical backdrop for the development of the model which described the correlations between the selected variables. Davis (1986) provided a basis for tracing the impact of external factors on internal beliefs, attitudes, and intentions. TAM was one of the most refutable models in information system literature (Dadayan and Ferro 2005). The model was used in different research settings such as online games (Hsu et al. 2004), online shopping (Vijayasararhy 2004), internet banking (Lai et al., 2005). To capture a person’s belief, Davis identified two antecedent variables or two particular beliefs that were determinants of intention. The two variables identified where perceived usefulness and perceived ease of use. Perceived usefulness (PU) was defined to capture the beliefs that using a specific application enabled a person to perform their job better. Perceived ease of use (PEOU) variable captured the belief that the target system was easy to use and was worth using. Davis suggested that PU was influenced by PEOU and that both predicted attitude (A) which defined a person’s intention to use the system. The relation between the variables was described in Figure 1.2. Davis (1989) suggested that all else being equal an application perceived to be easy to use than another was more likely to be accepted by users. Referring to Figure 1.2, the variable 1.2A influenced 1.2D which in turn influenced 1.2E. .2 A Perceived ease of use

Attitude towards use

1.2B Perceived ease of use 6

C

D Behavioural intention to use

E Actual system use

Prato CIRN-DIAC Community Informatics Conference 2010: PhD Stream

Figure 1.2 the Technology Acceptance Model, Davis (1989) While the model was parsimonious in nature the model was popular among researchers due to it empirical measurement scale. The new measurement scale for PU and PEOU had strong psychometric properties, and the relationship between self-reported measures of usage by the user proved to be empirically linked within the model. Davis believed that one of the most significant findings from his study was that usefulness was strongly linked to usage rather than ease of use. This view was also held by Hung-Pin Shih (2004) which was citied by Sandberg and Wahlberg (2006) who reported that PU had the greatest impact on an individual’s intention to use. Davis also suggested that from a causal perspective ease of use maybe an antecedent to usefulness, rather than a parallel or direct determinant of usage. Degree of Generality Mathieson (1991) in review of TAM suggested that Davis (1989) wanted to “use a belief set that ...readily generalizes to different computer systems and user populations”. Therefore, Davis believed that a person’s beliefs about usefulness and ease of use were the primary determinants of computer intention. His belief led to the generalizability of the model, which was achieved by its parsimony. Some researchers rejected this and attempted to evaluate or extend TAM (Venkatesh and Davis 2000) or integrate TAM with the others models (Davis, Bagozzi and Warshaw 1992). Wixom and Todd (2005) suggested that attempts to extend TAM have generally taken one of three approaches: by introducing factors from related models, by introducing additional or alternative belief factors, and by examining antecedents and moderators of perceived usefulness and perceived ease of use. Many researchers argued that the two variables proposed did not reflect objective reality. Social Variables TAM did not explicitly include any social variables. Many researchers argued that TAM was limited in determining computer adoption and therefore proposed the integration of more variables to capture human behaviour and social change (Davis and Venkatesh 2000; Venkatesh and Morris 2000;Carter and Belanger 2005). Legris et al (2003) emphasised the importance of social variables as they influenced a person’s attitude towards a technology as an individual’s belief could be over ruled by peers’ attitude towards the technology. Legris et al (2003) also argued that no external variables were defined in TAM and therefore self- reported use could be bias. He suggested a person could adopt a system by over rating the system performance and adopt a system that was not as functional. However, Davis (1989) postulated that external factors intervened indirectly and influenced PU and PEOU. Conversely, Davis and Venkatesh (2002) did introduce the social norm variable in TAM2.

References (Mathieson 1991) Adrian Sparkes, B. T. (2001). "The use of the Internet as a critical success factor for the marketing of Welsh agri-food SMEs in the twenty-first century." British Food Journal 103(5): 331-347. Ajzen, I. (2002). "Residual Effects of Past on Later Behavior: Habituation and Reasoned Action Perspectives." Personality and Social Psychology Review 6(2): 107. Ajzen, I. (2002). "Residual Effects of Past on Later Behaviour: Habituation and Reasoned Action Perspectives." Personality and Social Psychology Review, 6(2): 107-122. Davis, F. D. (1989). "Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology." MIS Quarterly 13(3): 319.

7

Prato CIRN-DIAC Community Informatics Conference 2010: PhD Stream Davis, F. D., R. P. Bagozzi, et al. (1989). "User acceptance of computer technology: a comparison of two theoretical models." Management Science 35(8): 982-1003. Department of Agriculture, F. a. F. (2008). Committee on the Uptake of Information Technology in Agriculture and in Rural Communities. F. a. F. Department of Agriculture. Dublin. 1: 44. Development, I. I. o. C. a. (2006). "ICTs for agricultural livelihoods Impact and lessons learned from IICD supported activities." Mathieson, K. (1991). "Predicting User Intentions: Comparing the Technology Acceptance Model with the Theory of Planned Behavior." Information Systems Research 2(3): p173-191. Ministry of Agriculture and Food, I. R. (2005). "The future of farming in Ireland." Eurochoices (Vol. 4) (No. 1) 6-11 6-11. Ogden, J. (2003). "Some problems with social cognition models: A pragmatic and conceptual analysis." HEALTH PSYCHOLOGY-HILLSDALE- 22(4): 424-428. Sandberg, K. W. and O. Wahlberg (2006). "Towards a Model of the Acceptance of Information and Communication Technology in Rural Small Businesses." Department of Information Technology and Media, Mid Sweden University, SE-851 70. van der Ploeg, J. D. (2000). "Revitalizing Agriculture: Farming Economically as Starting Ground for Rural Development." Sociologia Ruralis 40(4): 497-511. Warren, M. (2004). "Farmers online: drivers and impediments in adoption of Internet in UK agricultural businesses The Authors." Journal of Small Business and Enterprise Development 11(3): 371-381. Warren, M. F. (2002). "Adoption of ICT in agricultural management in the United Kingdom: the intra-rural digital divide." ZEMEDELSKA EKONOMIKA-PRAHA- 48(1): 1-8.

8